Investors in state bonds, beware. Four U.S. states — Illinois, New Jersey, Hawaii, and Connecticut — are sitting on time bombs. Between 35% and 51% of their annual revenues are likely to be needed to meet their total annual payment obligations on existing debt, retirement plans, and retiree healthcare. Yet the general obligation (GO) bonds of New Jersey, Hawaii, and Connecticut have a single A-rating from both Moody’s and S&P Global; the GO bonds of Illinois still have an investment-grade rating of triple-B.
Look at the table below, which compares two calculations of the current payment obligations for these four states as a percentage of state revenues — the first as reported by the states for fiscal year 2017; the second as calculated by Michael Cembalest and the Investment Strategy Team of JP Morgan Asset Management with more realistic assumptions. The table shows that the current payments of these four states are well-below a more realistic measure of their obligations to retired employees and bondholders.
Current payment obligations as a percentage of state revenues
State
As Reported by States
As Calculated with Realistic Assumptions
Illinois
26%
51%
New Jersey
17%
38%
Hawaii
21%
37%
Connecticut
22%
35%
Source: JP Morgan Asset Management, State Annual Financial Reports, Moody’s.
Let’s understand the main categories of state payment obligations and the unrealistic assumptions underlying the numbers reported by these states. Then let’s evaluate the actions that the newly elected officials of these states could reasonably take to substantially reduce the default risk on their GO bonds. In any event, let’s reject any efforts to get the federal government to bail out these states.
The largest state obligations stem from long-term liabilities to their defined benefit pension plans for current and retired state employees. State pension funds in 2016 reported total deficits of $1.4 trillion, with funding on average equal to just 66% of their reported obligations, according to the Pew Charitable Trust. A few states also have obligations to make contributions to defined contribution plans for state employees.
In assessing the present value of the unfunded obligations of their pension plans, we must choose an appropriate discount rate. The table assumes a discount rate of 6%. In pension math, the discount rate reflects the assumed investment returns on the state’s portfolio. The state’s current payment obligations are then derived by amortizing unfunded obligations over 30 years, assuming that both the existing assets and future contributions earn the assumed return.
While the discount rate used by state pension plans has been declining slowly to an average of 7.1%, it still is too high for a likely return on such a hypothetical portfolio. By contrast, the average discount rate used by corporate pension plans is 3.6% — the current return on a high-quality portfolio of corporate and government bonds over the next decade or so. We should accept 6% as a compromise favorable to the states — representing a likely long-term return on a diversified, low-cost portfolio comprised of 60% stocks and 40% bonds.
The table also uses a 6% discount rate to compute the present value of the unfunded liability of states for their retiree healthcare plans. These plans offer high-quality healthcare benefits to state employees and their families from their retirement to age 65 when they go on Medicare — and sometimes beyond age 65. This is often a lengthy period because many state employees retire at age 50 or 55.
The retiree healthcare plans of most states are heavily subsidized — with retirees paying low premiums and minimal deductibles for a broad range of medical services. Yet retiree healthcare plans on average have advance funding at an extremely low level — $46 billion in assets to meet $692 billion in obligations as of 2015, according to the Pew Charitable Trust.
The final category is the payments that states must make on their existing debts. These four states accounted for $152 billion in GO bonds in 2017 — 13% of GO bonds outstanding for all U.S. states. In this category, the table includes state payments due not only on GO bonds, but also on capital leases, lease revenue bonds, pension obligations bonds, and any negative balances in the state operating fund.
Reforms needed, but not taken
What can be done by Illinois, New Jersey, Hawaii, and Connecticut to manage down their long-term payment obligations? To totally close their funding gap by only one strategy, these four states would have to raise tax revenues by more than 12% or, alternatively, increase employee contributions by more than 400%, according to the authors of the table above. Although such measures would not be politically feasible, incremental movement in both areas would be needed as part of a general reform package.
Some states have started to offer less generous retirement plans for new employees — shifting partly or wholly to a defined contribution plan. Although these reforms take years to have a significant financial impact, they are important steps toward reducing future payment obligations of states. But such reforms have not yet been adopted in Illinois, New Jersey, Hawaii, and Connecticut.
Most state constitutions prevent any reductions in accrued benefits — for retirees and current employees. Yet the courts have split on the critical question of whether the benefit formulas for current employees may be changed for future years of state employment. Courts in Illinois, New Jersey, Hawaii, and Connecticut would seem receptive to certain changes in benefit formulas if their legislatures would apply such changes only to future pension accruals.
By contrast, the U.S. Supreme Court has decided that retiree healthcare benefits may generally be changed at the end of the collective bargaining agreement. This route is now legally permissible, subject to negotiations between states and their employees. As a result, quite a few states — but notably, not Illinois, New Jersey, Hawaii, and Connecticut — have substantially reduced their projected liabilities for retiree healthcare by increasing premium contributions and deductibles, as well as eliminating cost of living adjustments.
Although cities are typically allowed to file for bankruptcy, states may not. Accordingly, several commentators have suggested that Congress pass a special bankruptcy act for states, along the lines of the PROMESA legislation for Puerto Rico. Such legislation would allow states to abrogate their long-term obligations to bond holders and retired employees – leading to a complex negotiation among interested parties.
Such legislation should be strongly opposed. It would wreak havoc in the municipal bond market, undermine the valiant efforts of many states to manage down their long-term liabilities, and reward these four states that have overpromised state employee benefits and under-funded these promises.
In short, most states have been doing a reasonable job of limiting their payment obligations, while a few states have been irresponsible. Instead of bailing out these problem states, Congress should insist that they honor their past obligations to their bond holders and retired employees. At the same time, bond investors — together with credit rating agencies, accounting regulators and relevant local groups — should pressure Illinois, New Jersey, Hawaii, and Connecticut to promulgate financial reports based on more realistic assumptions and adopt bolder reforms reducing their future payment obligations.
Robert Pozen is a senior lecturer at MIT Sloan School of Management and a senior non-resident fellow at the Brookings Institution.
